The present invention relates to antibodies and fragments thereof which can be used to inhibit activity of metalloproteins, such as metalloproteases, and to methods which utilize same for treating diseases such as metastatic cancer which are associated with abnormal activity of a metalloprotein.
The matrix metalloproteins (MMPs) are key enzymes participating in remodeling of the extracellular matrix (ECM). These enzymes are capable of destroying a variety of connective tissue components of articular cartilage or basement membranes.
The human MMP gene family consists of at least 28 structurally related proteins (see FIG. 1), which share a similar overall spherical topology (FIG. 2 and Borkakoti, 1998). Each MMP is secreted as an inactive, latent pro-enzyme. The catalytic zinc domain is composed of about 180 amino acids wherein the highly conserved sequence HE-GH-LGL-H provides the three histidine (i.e., H) residues which bind to the metal Zn(2+) ion. The forth-binding site of the catalytic zinc ion in the pro-enzyme is bound to a cystein residue (Morgunova et al., 1999), which upon enzyme activation dissociates from the active site (Van Wart and Birkedal-Hansen, 1990). As a result, the forth-binding site in the activated MMPs is taken up by a water molecule, which is also hydrogen-bonded to a conserved glutamic residue. This process facilitates the hydrolysis of a peptide bond of the target substrate with the activated water molecule.
The uncontrolled breakdown of connective tissue by metalloproteases is a feature of many pathological conditions, probably resulting from an excess of MMP activity or from an imbalanced ratio between the natural MMP tissue inhibitors (TIMPs) and MMPs. TIMPs inhibit MMPs by forming stoichiometric complexes with the active zinc binding site of MMPs (Gomez et al., 1997; Henriet at al., 1999; Bode et al., 1999; Will et al., 1996). When TIMPs levels are insufficient, a progressive slow degradation of the ECM may lead to loss of cartilage matrix in rheumatoid arthritis (Walakovits et al., Arthritis Rheum, 35:35-42, 1992) and osteoarthritis (Dean et al., J. Clin. Invest. 84:678-685, 1989) or bone matrix degradation in osteoporosis (Hill et al., Biochem. J. 308: 167-175, 1995). In other situations, such as congestive heart failure, rapid degradation of the heart's ECM may occur (Armstrong et al., Canadian J. Cardiol. 10: 214-220, 1994).
Other pathological condition, which are also related to unregulated activity of MMPs, include the rapid remodeling of the ECM by metastatic tumor cells. In such conditions the activated MMPs are either expressed by the cancer cells or by the surrounding tissues. There is considerable evidence that MMPs are involved in the growth and spread of tumors (e.g., see Davidson et al., Chemistry & Industry, 258-261, 1997, and references therein). In the process of tumor metastasis, MMPs are used to break down the ECM, allowing primary tumor cancer cells to invade neighboring blood vessels where they are transported to different organs and establish secondary tumors. The invasive growth at these secondary sites is mediated by MMPs, which break down the tissue. In addition, MMP activity contributes to the invasive in-growth of new blood vessels, also termed angiogenesis, which is required for tumors to grow above a certain size.
Given the broad role of MMPs in human physiology and pathology, it is not surprising that numerous efforts have been affected to design drugs, which inhibit MMP excessive activity.
Drug discovery efforts have focused on inhibitor classes that contain a functional group which coordinates the zinc ion to thereby inactivate the target MMP. One such inhibitor class is the hydroxamate inhibitors, small peptide analogs of fibrillar collagens, which specifically interact in a bidentate manner via the hydroxyl and carbonyl oxygens of the hydroxamic group with the zinc ion in the catalytic site [Grams et al., (1995), Biochem. 34: 14012-14020; Bode et al., (1994), EMBO J., 13: 1263-1269].
Hydroxamate-based MMP inhibitors are usually composed of either a carbon back-bone (WO 95/29892, WO 97/24117, WO 97/49679 and EP 0780386), a peptidyl back-bone (WO 90/05719, WO 93/20047, WO 95/09841 and WO 96/06074) or a peptidomimetic back-bone [Schwartz et al., Progr. Med. Chem., 29: 271-334 (1992); Rasmussen et al., Pharmacol. Ther., 75: 69-75 (1997); Denis et al., Invest. New Drugs, 15: 175-185 (1997)]. Alternatively, they contain a sulfonamido sulfonyl group which is bonded on one side to a phenyl ring and a sulfonamido nitrogen which is bonded to an hydroxamate group via a chain of one to four carbon atoms (EP 0757984 A1).
Other peptide-based MMP inhibitors are thiol amides which exhibit collagenase inhibition activity (U.S. Pat. No. 4,595,700), N-carboxyalkyl derivatives containing a biphenylethylglycine which inhibit MMP-3, MMP-2 and collagenase (Durette, et al., WO-9529689), lactam derivatives which inhibit MMPs, TNF-alpha and aggrecanase (see U.S. Pat. No. 6,495,699) and Tricyclic sulfonamide compounds (see U.S. Pat. No. 6,492,422).
Although peptide-based MMP inhibitors have a clear therapeutic potential their use in clinical therapy is limited. Peptide-based hydroxamate are costly to produce and have low metabolic stability and oral bioavailability [e.g., batimastat (BB-94)]. These compounds are rapidly glucuronidated, oxidized to carboxylic acid and excreted in the bile [Singh et al., Bioorg. Med. Chem. Lett. 5: 337-342, 1995; Hodgson, “Remodelling MMPIs”, Biotechnology 13: 554-557, 1995)]. In addition, peptide-based MMP inhibitors often exhibit the same or similar inhibitory effects against each of the MMP enzymes. For example, batimastat is reported to exhibit IC50 values of about 1 to about 20 nM against each of MMP-1, MMP-2, MMP-3, MMP-7, and MMP-9 [Rasmussen et al., Pharmacol. Ther., 75(1): 69-75 (1997)]. Furthermore, the use of several hydroxamate inhibitors was associated with severe side effects such as muscoloskeletal problems with marimastat (BB-2516), widespread maculopapular rash with CGS27023A (Novartis) [Levitt et al., 2001, Clin. Cancer Res. 7: 1912-1922] and liver abnormalities, anemia, shoulder and back pain, thrombocytopenia, nausea, fatigue, diarrhea and deep vein thrombosis with BAY12-9566 (Bayer) [Heath et al., 2001, Cancer Chemother. Pharmacol. 48: 269-274]. Moreover, phase III clinical trials on advanced cancer patients with marimastat, prinomastat (AG 3340, Agouron) and Bay 12-9566 demonstrated no clinical efficacy in inhibiting metastasis (Zucker et al., 2000, Oncogene 19: 6642-50).
Other MMP inhibitors are the chemically modified nonmicrobial tetracyclines (CMTs) that were shown to block expression of several MMPs in vitro. However, in vivo efficacy of these compounds was found to be limited, e.g., the CMT inhibitor, doxycycline, reduced tissue levels of MMP-1 but not MMP-2, 3, or −9 in atherosclerotic carotid plaques in human patients (Axisa et al., 2002, Stroke 33: 2858-2864).
Recently, a mechanism-based MMP inhibitor, SB-3CT, was designed according to the X-ray crystallographic information of the MMP active site (Brown et al., 2000). X-ray absorption studies revealed that binding of this molecule to the catalytic zinc reconstructs the conformational environment around the active site metal ion back to that of the pro-enzyme [Kleifeld et al., 2001, J. Biol. Chem. 276: 17125-31]. However, the therapeutic efficacy obtained with this agent is yet to be determined.
Another class of natural inhibitors is monoclonal antibodies. Several antibodies have been raised against specific peptide sequences within the catalytic domain MMP-1 (Galvez et al., 2001, J. Biol. Chem., 276: 37491-37500). However, although these antibodies could inhibit the in-vitro activity of MMP, results demonstrating the in-vivo effectiveness of such antibodies have not been demonstrated.
As described hereinabove, the catalytic site of MMPs includes a coordinated metal ion which becomes available for substrate binding following enzyme activation (see FIGS. 2a-c). It is thus conceivable that conventional antibodies directed at the primary amino acid sequence of the enzyme would not distinguish the active form from the inactive form of the enzyme and hence would not serve as potent inhibitors of such enzymes.
While reducing the present invention to practice, the present inventors have uncovered that in sharp contrast to the above, antibodies which recognize electronic and structural determinants of the catalytic site of MMPs are potent inhibitors thereof and as such can be used to treat diseases associated with imbalanced MMP activity.